fortio update #460

ldemailly · 2017-07-09T03:07:44Z

build and use fortio in perf/standalone scripts
fix for having more connections than requested
disabling compression by default (saves cpu), re-enable using new -compression flag
updating sample in README.md to a test using echosrv (slightly lower latency)
(misc/minor) also changing threshold for sleep time histogram to 5%

- build and use fortio in perf/standalone scripts - (win) fix for having more connections than requested

Also changing threshold for sleep time histogram to 5%

Now getting exactly the expected number of sockets But performance is still significantly lower than wrk

ldemailly · 2017-07-15T00:30:38Z

this fixes 2 entries from #453

ldemailly · 2017-07-15T02:15:52Z

this version now uses the exact/expected number of sockets

it still uses more cpu than wrk but it can now trigger 60k qps (up from 40 but still lower than wrk's 100k qps) out of echosrv. the profile isn't clear

44.44s of 47.82s total (92.93%)
Dropped 287 nodes (cum <= 0.24s)
Showing top 10 nodes out of 72 (cum >= 0.25s)
      flat  flat%   sum%        cum   cum%
    27.61s 57.74% 57.74%     27.76s 58.05%  syscall.Syscall
     7.37s 15.41% 73.15%      7.37s 15.41%  runtime.kevent
     3.06s  6.40% 79.55%      3.06s  6.40%  runtime.usleep
     2.32s  4.85% 84.40%      2.32s  4.85%  runtime.mach_semaphore_wait
     2.06s  4.31% 88.71%      2.06s  4.31%  runtime.mach_semaphore_signal
     1.15s  2.40% 91.11%      1.19s  2.49%  runtime.freedefer
     0.47s  0.98% 92.10%      1.23s  2.57%  runtime.selectgoImpl
     0.18s  0.38% 92.47%      0.55s  1.15%  net/http.setRequestCancel.func3
     0.13s  0.27% 92.74%      0.60s  1.25%  runtime.mallocgc
     0.09s  0.19% 92.93%      0.25s  0.52%  runtime.lock

Probably some locking is getting in the way (I tried to avoid all locking by having separate objects in each goroutine but the http client still seems to shares and lock things)

ldemailly · 2017-07-16T07:58:12Z

With #470 it's now very close to wrk

ran gazelle

douglas-reid · 2017-07-20T18:11:09Z

devel/fortio/http.go

+		url,
+		req,
+		&http.Client{
+			Timeout: 3 * time.Second,


curious: any special reason for 3s timeouts ?

it means (I think) the max duration will be 3000ms before giving up - given we're expecting data in the single digit ms range most of the time, it seemed enough but I don't really have a good reason - probably should make it configurable or not set it

douglas-reid · 2017-07-20T18:11:45Z

devel/fortio/periodic.go

@@ -170,7 +170,7 @@ func (r *periodicRunner) Run() {
 	fmt.Printf("Ended after %v : %d calls. qps=%.5g\n", elapsed, functionDuration.Count, actualQPS)
 	if useQPS {
 		percentNegative := 100. * float64(sleepTime.hdata[0]) / float64(sleepTime.Count)
-		if percentNegative > 3 {
+		if percentNegative > 5 {


worth documenting choice of 5 here?

I'll add a comment, thx, I think it's somewhat arbitrary - it's how often we are falling behind for the target qps - 5% is in the noise (actually maybe 10) - the only effect is more verbose debugging and the warning

istio-testing · 2017-07-20T18:45:31Z

Jenkins job istio/presubmit passed

And misc comment/leftover from #460 etc

* fortio update - build and use fortio in perf/standalone scripts - fix for having more connections than requested * Updating sample to test using echosrv (slightly lower latency) Also changing threshold for sleep time histogram to 5% * Adding -gomaxprocs flag * One client per goroutine Now getting exactly the expected number of sockets But performance is still significantly lower than wrk * Adding profiler option * Homegrown http client On my Mac: Old net/http : 19.7k qps with 2 threads (-http -1) New -http 1 : 30.5k ops with 2 threads And 1/8th of the user cpu Interface to select between the 2 clients * More comments and backtrack test * http_test was missing from bazel ran gazelle * Added benchmark and fix folding bugs Folding is 4x faster than built in, 1/3rd of allocs Search is 10x faster, no alloc instead of a lot ``` BenchmarkASCIIFoldNormalToLower-8 3000000 416 ns/op 96 B/op 3 allocs/op BenchmarkASCIIFoldCustomToLowerMap-8 5000000 371 ns/op 96 B/op 3 allocs/op BenchmarkASCIIToUpper-8 20000000 104 ns/op 32 B/op 1 allocs/op BenchmarkFoldFind0-8 200000 7621 ns/op 2112 B/op 3 allocs/op BenchmarkFoldFind-8 2000000 669 ns/op 0 B/op 0 allocs/op ``` * chunked decoding unit test for parsing, reads the first chunk * fasthttp client wrapper (-http 3) - added some doc in WORKSPACE about wtool - initial support for fasthttp * Make linter happy though I might drop fasthttp as it’s not faster (but it’s more complete) and adds a lot of dependencies * Split connect out * Fixing merge error * adding fortio Logger * Fixing bazel build And misc comment/leftover from #460 etc * Shorter logger names * Switched to new logging * Eradicate Verbosity leftovers * Added -logprefix and -logcaller; Refactor start on http.go also fixed a couple of calls missing % codes * Minor edits, saving to GitHub wip * Woot http1.1 works now Still unwieldy but working ! * Some profiler optimizations * More profiler opt * Make lint happy by making checking connection closed header a flag * Dropping fasthttp dependency * Moved 3 changes to #495 * Minor: use version in echoers, use 3 digits * Code review updates (thx doug!) + serious chunked bug fix Wasn’t working with last chunk (0\r\n\r\n) being in a separate frame * Code review comment thx doug!

…#460) * Implement validation using descriptors for the metrics aspect. * Add back in default for latency expr Former-commit-id: f229ffdf83e68abaf35666e365988e1ebced060f

* fortio update - build and use fortio in perf/standalone scripts - fix for having more connections than requested * Updating sample to test using echosrv (slightly lower latency) Also changing threshold for sleep time histogram to 5% * Adding -gomaxprocs flag * One client per goroutine Now getting exactly the expected number of sockets * Adding profiler option * http_test was missing from bazel ran gazelle Former-commit-id: 01fadb6

* fortio update - build and use fortio in perf/standalone scripts - fix for having more connections than requested * Updating sample to test using echosrv (slightly lower latency) Also changing threshold for sleep time histogram to 5% * Adding -gomaxprocs flag * One client per goroutine Now getting exactly the expected number of sockets But performance is still significantly lower than wrk * Adding profiler option * Homegrown http client On my Mac: Old net/http : 19.7k qps with 2 threads (-http -1) New -http 1 : 30.5k ops with 2 threads And 1/8th of the user cpu Interface to select between the 2 clients * More comments and backtrack test * http_test was missing from bazel ran gazelle * Added benchmark and fix folding bugs Folding is 4x faster than built in, 1/3rd of allocs Search is 10x faster, no alloc instead of a lot ``` BenchmarkASCIIFoldNormalToLower-8 3000000 416 ns/op 96 B/op 3 allocs/op BenchmarkASCIIFoldCustomToLowerMap-8 5000000 371 ns/op 96 B/op 3 allocs/op BenchmarkASCIIToUpper-8 20000000 104 ns/op 32 B/op 1 allocs/op BenchmarkFoldFind0-8 200000 7621 ns/op 2112 B/op 3 allocs/op BenchmarkFoldFind-8 2000000 669 ns/op 0 B/op 0 allocs/op ``` * chunked decoding unit test for parsing, reads the first chunk * fasthttp client wrapper (-http 3) - added some doc in WORKSPACE about wtool - initial support for fasthttp * Make linter happy though I might drop fasthttp as it’s not faster (but it’s more complete) and adds a lot of dependencies * Split connect out * Fixing merge error * adding fortio Logger * Fixing bazel build And misc comment/leftover from #460 etc * Shorter logger names * Switched to new logging * Eradicate Verbosity leftovers * Added -logprefix and -logcaller; Refactor start on http.go also fixed a couple of calls missing % codes * Minor edits, saving to GitHub wip * Woot http1.1 works now Still unwieldy but working ! * Some profiler optimizations * More profiler opt * Make lint happy by making checking connection closed header a flag * Dropping fasthttp dependency * Moved 3 changes to #495 * Minor: use version in echoers, use 3 digits * Code review updates (thx doug!) + serious chunked bug fix Wasn’t working with last chunk (0\r\n\r\n) being in a separate frame * Code review comment thx doug! Former-commit-id: e09dcaf

* Implement validation using descriptors for the metrics aspect. * Add back in default for latency expr Former-commit-id: 28b5ce56493e385b065f1e7c50d7cab335bae37c

* fortio update - build and use fortio in perf/standalone scripts - fix for having more connections than requested * Updating sample to test using echosrv (slightly lower latency) Also changing threshold for sleep time histogram to 5% * Adding -gomaxprocs flag * One client per goroutine Now getting exactly the expected number of sockets * Adding profiler option * http_test was missing from bazel ran gazelle Former-commit-id: 01fadb6

* fortio update - build and use fortio in perf/standalone scripts - fix for having more connections than requested * Updating sample to test using echosrv (slightly lower latency) Also changing threshold for sleep time histogram to 5% * Adding -gomaxprocs flag * One client per goroutine Now getting exactly the expected number of sockets But performance is still significantly lower than wrk * Adding profiler option * Homegrown http client On my Mac: Old net/http : 19.7k qps with 2 threads (-http -1) New -http 1 : 30.5k ops with 2 threads And 1/8th of the user cpu Interface to select between the 2 clients * More comments and backtrack test * http_test was missing from bazel ran gazelle * Added benchmark and fix folding bugs Folding is 4x faster than built in, 1/3rd of allocs Search is 10x faster, no alloc instead of a lot ``` BenchmarkASCIIFoldNormalToLower-8 3000000 416 ns/op 96 B/op 3 allocs/op BenchmarkASCIIFoldCustomToLowerMap-8 5000000 371 ns/op 96 B/op 3 allocs/op BenchmarkASCIIToUpper-8 20000000 104 ns/op 32 B/op 1 allocs/op BenchmarkFoldFind0-8 200000 7621 ns/op 2112 B/op 3 allocs/op BenchmarkFoldFind-8 2000000 669 ns/op 0 B/op 0 allocs/op ``` * chunked decoding unit test for parsing, reads the first chunk * fasthttp client wrapper (-http 3) - added some doc in WORKSPACE about wtool - initial support for fasthttp * Make linter happy though I might drop fasthttp as it’s not faster (but it’s more complete) and adds a lot of dependencies * Split connect out * Fixing merge error * adding fortio Logger * Fixing bazel build And misc comment/leftover from istio#460 etc * Shorter logger names * Switched to new logging * Eradicate Verbosity leftovers * Added -logprefix and -logcaller; Refactor start on http.go also fixed a couple of calls missing % codes * Minor edits, saving to GitHub wip * Woot http1.1 works now Still unwieldy but working ! * Some profiler optimizations * More profiler opt * Make lint happy by making checking connection closed header a flag * Dropping fasthttp dependency * Moved 3 changes to istio#495 * Minor: use version in echoers, use 3 digits * Code review updates (thx doug!) + serious chunked bug fix Wasn’t working with last chunk (0\r\n\r\n) being in a separate frame * Code review comment thx doug! Former-commit-id: e09dcaf

* Default to POD_NAMESPACE for -n flag * Fix bad merge

* fortio update - build and use fortio in perf/standalone scripts - fix for having more connections than requested * Updating sample to test using echosrv (slightly lower latency) Also changing threshold for sleep time histogram to 5% * Adding -gomaxprocs flag * One client per goroutine Now getting exactly the expected number of sockets * Adding profiler option * http_test was missing from bazel ran gazelle Former-commit-id: 01fadb6

* fortio update - build and use fortio in perf/standalone scripts - fix for having more connections than requested * Updating sample to test using echosrv (slightly lower latency) Also changing threshold for sleep time histogram to 5% * Adding -gomaxprocs flag * One client per goroutine Now getting exactly the expected number of sockets But performance is still significantly lower than wrk * Adding profiler option * Homegrown http client On my Mac: Old net/http : 19.7k qps with 2 threads (-http -1) New -http 1 : 30.5k ops with 2 threads And 1/8th of the user cpu Interface to select between the 2 clients * More comments and backtrack test * http_test was missing from bazel ran gazelle * Added benchmark and fix folding bugs Folding is 4x faster than built in, 1/3rd of allocs Search is 10x faster, no alloc instead of a lot ``` BenchmarkASCIIFoldNormalToLower-8 3000000 416 ns/op 96 B/op 3 allocs/op BenchmarkASCIIFoldCustomToLowerMap-8 5000000 371 ns/op 96 B/op 3 allocs/op BenchmarkASCIIToUpper-8 20000000 104 ns/op 32 B/op 1 allocs/op BenchmarkFoldFind0-8 200000 7621 ns/op 2112 B/op 3 allocs/op BenchmarkFoldFind-8 2000000 669 ns/op 0 B/op 0 allocs/op ``` * chunked decoding unit test for parsing, reads the first chunk * fasthttp client wrapper (-http 3) - added some doc in WORKSPACE about wtool - initial support for fasthttp * Make linter happy though I might drop fasthttp as it’s not faster (but it’s more complete) and adds a lot of dependencies * Split connect out * Fixing merge error * adding fortio Logger * Fixing bazel build And misc comment/leftover from #460 etc * Shorter logger names * Switched to new logging * Eradicate Verbosity leftovers * Added -logprefix and -logcaller; Refactor start on http.go also fixed a couple of calls missing % codes * Minor edits, saving to GitHub wip * Woot http1.1 works now Still unwieldy but working ! * Some profiler optimizations * More profiler opt * Make lint happy by making checking connection closed header a flag * Dropping fasthttp dependency * Moved 3 changes to #495 * Minor: use version in echoers, use 3 digits * Code review updates (thx doug!) + serious chunked bug fix Wasn’t working with last chunk (0\r\n\r\n) being in a separate frame * Code review comment thx doug! Former-commit-id: e09dcaf

* Update global dictionary for new attributes. * Move new attributes to the bottom of the file for backward compatible.

Co-authored-by: maistra-bot <null>

ldemailly added 2 commits July 8, 2017 19:03

fortio update

d1b2134

- build and use fortio in perf/standalone scripts - (win) fix for having more connections than requested

Updating sample to test using echosrv (slightly lower latency)

6707ad8

Also changing threshold for sleep time histogram to 5%

googlebot added the cla: yes label Jul 9, 2017

ldemailly added 3 commits July 14, 2017 14:31

Merge branch 'master' into ldemailly-bench6

993245e

Adding -gomaxprocs flag

e3a49a7

One client per goroutine

784f9bc

Now getting exactly the expected number of sockets But performance is still significantly lower than wrk

ldemailly added 2 commits July 14, 2017 19:10

Adding profiler option

88205e2

Merge branch 'master' into ldemailly-bench6

3052cdd

istio deleted a comment from istio-testing Jul 15, 2017

ldemailly requested a review from douglas-reid July 15, 2017 02:16

istio deleted a comment from istio-testing Jul 15, 2017

ldemailly mentioned this pull request Jul 16, 2017

fortio: optimized http client, simplified logging #470

Merged

ldemailly and others added 2 commits July 16, 2017 01:26

http_test was missing from bazel

5372781

ran gazelle

Merge branch 'master' into ldemailly-bench6

da5c613

istio deleted a comment from istio-testing Jul 16, 2017

douglas-reid approved these changes Jul 20, 2017

View reviewed changes

Merge branch 'master' into ldemailly-bench6

3730839

ldemailly merged commit 01fadb6 into master Jul 20, 2017

ldemailly deleted the ldemailly-bench6 branch July 20, 2017 21:01

ldemailly added a commit that referenced this pull request Jul 21, 2017

Fixing bagel build

1d33ae9

And misc comment/leftover from #460 etc

mandarjog pushed a commit that referenced this pull request Oct 31, 2017

Default to POD_NAMESPACE for -n flag (#460)

0fb83dd

* Default to POD_NAMESPACE for -n flag * Fix bad merge

guptasu pushed a commit to guptasu/istio that referenced this pull request Jun 11, 2018

Update global dictionary for new attributes. (istio#460)

903cfe9

* Update global dictionary for new attributes. * Move new attributes to the bottom of the file for backward compatible.

kyessenov pushed a commit to kyessenov/istio that referenced this pull request Aug 13, 2018

Fix a tcp filter crash. (istio#460)

4b1b15c

luksa pushed a commit to luksa/istio that referenced this pull request Sep 20, 2022

Automator: Update proxy (istio#460)

1f2e152

Co-authored-by: maistra-bot <null>

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

fortio update #460

fortio update #460

ldemailly commented Jul 9, 2017 •

edited

ldemailly commented Jul 15, 2017

ldemailly commented Jul 15, 2017 •

edited

ldemailly commented Jul 16, 2017 •

edited

douglas-reid Jul 20, 2017

ldemailly Jul 20, 2017

douglas-reid Jul 20, 2017

ldemailly Jul 20, 2017

istio-testing commented Jul 20, 2017

fortio update #460

fortio update #460

Conversation

ldemailly commented Jul 9, 2017 • edited

ldemailly commented Jul 15, 2017

ldemailly commented Jul 15, 2017 • edited

ldemailly commented Jul 16, 2017 • edited

douglas-reid Jul 20, 2017

Choose a reason for hiding this comment

ldemailly Jul 20, 2017

Choose a reason for hiding this comment

douglas-reid Jul 20, 2017

Choose a reason for hiding this comment

ldemailly Jul 20, 2017

Choose a reason for hiding this comment

istio-testing commented Jul 20, 2017

ldemailly commented Jul 9, 2017 •

edited

ldemailly commented Jul 15, 2017 •

edited

ldemailly commented Jul 16, 2017 •

edited